[fpc-pascal] Re: Porting Discussion

[Marco van de Voort]

> On 24 Jun 2008, at 15:02, Marco van de Voort wrote:

> That's a different matter and may actually be a real reason to invest  
> in Cocoa/Obj-C, but for most apps 64 bit needs are still quite far  
> off. This can also happen gradually since you can perfectly combine  
> Cocoa and Carbon in a single app (they're also adding functionality to  
> make that easier).

Nobody likes to invest in a trajectory that is uncertain. 
 
> >> Nobody has asked for this. But flooding the Mac market with
> >> quick&dirty Windows ports is going to be as commercially successful  
> >> as
> >> Kylix was on Linux,
> >
> > I do not agree with the analogy. The problem with Kylix was that it  
> > targeted
> > a (not yet) existing market,
> 
> And because the IDE was slow, buggy and crashing all the time to the  
> point where it was virtually unusable (at least that's what I remember  
> from Michael's mails). 

Compared to Delphi standards it was bad. Of course since pretty much the
only other choice was either primordial Eclipse or commandline, it was above
its direct competition in every way, despite of the problems.

> The reason (or at least one of the reasons): it  
> was a quick&dirty Windows port using an early version of Wine rather  
> than a native Linux port.

Winelib, and only the earlier versions iirc.
 
> > So you seem to miss my point. Namely that for most of the porters a
> > proper budget to do it right is simply missing.
> 
> I'm very well aware of that, I'm just saying what the consequences are/ 
> will be of that.

We'll see. Like the other commenters on the thread said, I also have the
feeling that this is mostly the formal stand of the Mac innercrowd, echoed
over various macfanatics websites.  

I've no way to prove that point however, so I haven't brought it up before. 

Fact is however that the center Mac will become increasingly outside the
former strongholds. (I even saw FreeBSD core developers with macbooks on
FOSDEM. And no, it wasn't running FreeBSD) 

> > So it's either "quick&dirty" or not. Both for the developers, as for the
> > users. And IMHO it is arrogant to state that then there can be no such
> > apps at all.
> 
> Who said there cannot be such apps? I'm only saying that in general it  
> is a bad idea to create them because they won't be received very well  
> by the Mac users/market at large (and if someone else creates one with  
> a proper interface which does more or less the same as yours, your  
> small investment will probably soon enough be largely in vain, except  
> possibly for those helpdesks that only want to shut up annoying users).

Well, the point is that you don't sell to Macers at all. You sell to people
that have to support Mac, and often grudgingly. They don't have a Calimero
complex, and are not indoctrinated by Mac sites to have unrealistic
expectations.

Most central IT organizations are not Mac oriented, and the standards for
specialist custom software are generally much lower than for shrinkwrap
software (excluding some of the bulksoftware on Windows. Whoever came up
with the idea that something that resembled a blob if spilt paint is a
perfect UI should be shot). Money is much harder to make in shrinkwrap
software anyway.

> It's just market advice, not a moral judgement.

If it is really about real life  markets, then we should first discuss
distribution channels. Selling is a lot less about technical details then
most people think. Which is why big companies piece of the bulk shrinkwrap
software market gets bigger every year.